Effect of Intraoperative Arterial Hypotension on the Risk of Perioperative Stroke After Noncardiac Surgery: A Retrospective Multicenter Cohort Study

Abstract

BACKGROUND: 

Intraoperative cerebral blood flow is mainly determined by cerebral perfusion pressure and cerebral autoregulation of vasomotor tone. About 1% of patients undergoing noncardiac surgery develop ischemic stroke. We hypothesized that intraoperative hypotension within a range frequently observed in clinical practice is associated with an increased risk of perioperative ischemic stroke within 7 days after surgery.

METHODS: 

Adult noncardiac surgical patients undergoing general anesthesia at Beth Israel Deaconess Medical Center and Massachusetts General Hospital between 2005 and 2017 were included in this retrospective cohort study. The primary exposure was intraoperative hypotension, defined as a decrease in mean arterial pressure (MAP) below 55 mm Hg, categorized into no intraoperative hypotension, short (<15 minutes, median [interquartile range {IQR}], 2 minutes [1–5 minutes]) and prolonged (≥15 minutes, median [IQR], 21 minutes [17–31 minutes]) durations. The primary outcome was a new diagnosis of early perioperative ischemic stroke within 7 days after surgery. In secondary analyses, we assessed the effect of a MAP decrease by >30% from baseline on perioperative stroke. Analyses were adjusted for the preoperative STRoke After Surgery (STRAS) prediction score, work relative value units, and duration of surgery.

RESULTS: 

Among 358,391 included patients, a total of 1553 (0.4%) experienced an early perioperative ischemic stroke. About 42% and 3% of patients had a MAP of below 55 mm Hg for a short and a prolonged duration, and 49% and 29% had a MAP decrease by >30% from baseline for a short and a prolonged duration, respectively. In an adjusted analysis, neither a MAP <55 mm Hg (short duration: adjusted odds ratio [OR_adj], 0.95; 95% confidence interval [CI], 0.85–1.07; P = .417 and prolonged duration: OR_adj, 1.18; 95% CI, 0.91–1.55; P = .220) nor a MAP decrease >30% (short duration: OR_adj, 0.97; 95% CI, 0.67–1.42; P = .883 and prolonged duration: OR_adj, 1.30; 95% CI, 0.89–1.90; P = .176) was associated with early perioperative stroke. A high a priori stroke risk quantified based on preoperatively available risk factors (STRAS prediction score) was associated with longer intraoperative hypotension (adjusted incidence rate ratio, 1.04; 95% CI, 1.04–1.05; P < .001 per 5 points of the STRAS prediction score).

CONCLUSIONS: 

This study found no evidence to conclude that intraoperative hypotension within the range studied was associated with early perioperative stroke within 7 days after surgery. These findings emphasize the importance of perioperative cerebral blood flow autoregulation to prevent ischemic stroke.

KEY POINTS

• Question: Is intraoperative hypotension associated with an ischemic stroke within 7 days after surgery?
• Findings: In this retrospective hospital registry study of >350,000 noncardiac surgical patients, after adjusting for preoperative risk factors and markers of procedural severity, neither an intraoperative mean arterial pressure of <55 mm Hg nor a decrease by >30% from baseline was associated with early perioperative ischemic stroke.
• Meaning: Cerebral autoregulation may play an essential role in maintaining cerebral blood flow to prevent perioperative ischemic stroke within the range of hypotension investigated in this study.

Approximately 313 million operative procedures were performed worldwide in 2015.¹ It has further been estimated that 4.3 million people worldwide die within 30 days of surgery each year.² Perioperative stroke is the most severe neurological complication after surgery, since it can be fatal or can cause severe complications.^3,4 The estimated stroke incidence in patients undergoing noncardiac surgery varies from 5% after carotid endarterectomy to <1% after general surgery, but the actual incidence is likely higher.^3,5

In 2017, the evidence-based hypertension clinical practice guidelines recommended intensive blood pressure control for primary and secondary stroke prevention. These recommendations were supported by substantial evidence that blood pressure management to a target goal of <130/80 mm Hg minimized the risk of stroke.^6,7 However, acutely decreasing blood pressure has been demonstrated to cause signs of cerebral ischemia,^8,9 which should lead clinicians to be cautious about intraoperative blood pressure decreases. Available literature differs significantly in the definition and role of intraoperative hypotension as an independent risk factor of perioperative stroke.^10,11 While some reports showed that decreasing blood pressure increases the risk of perioperative ischemic stroke,^12,13 others reported no association.^14–16

In this retrospective multicenter cohort study, we hypothesized that intraoperative hypotension is associated with an increased risk of perioperative ischemic stroke within 7 days after surgery.

METHODS

Study Design and Data Collection

This retrospective hospital registry study analyzed surgical cases performed between November 2005 and September 2017 at Beth Israel Deaconess Medical Center (BIDMC, Boston, MA) and between January 2007 and December 2015 at Massachusetts General Hospital (MGH, Boston, MA). Data obtained from clinical databases at MGH and BIDMC were deidentified and subsequently combined. Details on establishing a data repository are provided in Supplemental Digital Content 1, Appendix S1.1, https://links.lww.com/AA/D549. The Committee on Clinical Investigations affiliated with BIDMC (protocol number: 2019P000442) approved the study with a waiver of informed consent. A reliance agreement was established with the Partners Human Research Committee at MGH (Streamlined, Multisite, Accelerated Resources for Trials [SMART] institutional review board [IRB], request number 2070). This article adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational research (Supplemental Digital Content 2, https://links.lww.com/AA/D550). Patients or the public were not involved in the design or conduct of this study.

Study Population

Adult patients undergoing noncardiac surgery under general anesthesia were eligible for inclusion in this study. Patients with an American Society of Anesthesiologists (ASA) physical status classification of VI (brain death) or with missing covariate data were excluded.

Intraoperative Hypotension

Intraoperative hypotension was defined a priori as the duration of a mean arterial pressure (MAP) of <55 mm Hg. At a reviewers’ suggestion, we further evaluated the effect of a decrease of MAP by >30% from baseline on early perioperative ischemic stroke within 7 days in a subgroup of patients with available data.¹³ The baseline blood pressure was defined as the mean arterial blood pressure measured during preoperative evaluation. If these data were unavailable, the blood pressure from the preoperative holding area or the preinduction values were used. We removed artifacts, values out of range, and observations with an abrupt change in values by using a previously published data cleaning process, as described in detail in Supplemental Digital Content 1, Appendix S1.2, https://links.lww.com/AA/D549.¹⁷ Given the controversy and wide range of definitions of intraoperative hypotension throughout available literature,^10,11,18 we categorized the exposure variable into groups of no intraoperative hypotension as well as a short (<15 minutes) and a prolonged (≥15 minutes) duration of intraoperative hypotension.

Early Perioperative Ischemic Stroke

The primary outcome, early perioperative ischemic stroke, was defined as a new diagnosis of acute ischemic perioperative stroke occurring within 7 days after surgery and billed within the respective health care network.¹⁹ The respective diagnoses were identified through International Classification of Diseases, Ninth/Tenth Revision (ICD-9/10), diagnostic codes (Supplemental Digital Content 1, Table S1, https://links.lww.com/AA/D549). Additionally, the diagnosis of perioperative stroke was validated by medical record review performed by an interdisciplinary team of research fellows and a neurologist as previously described by our group (Supplemental Digital Content 1, Appendix S1.3, https://links.lww.com/AA/D549).^20,21

Covariate Model

The covariate model was established a priori utilizing a model allowing for preoperative assessment of a patient’s perioperative ischemic stroke risk (STRoke After Surgery [STRAS] prediction score) developed by our group.²³ Subsequently, we added intraoperative risk determinants of surgical complexity (duration of surgery and work relative value units [workRVUs]) to the primary covariate model and additional variables in sensitivity analyses. Details on the STRAS prediction score and categorization of covariates used in the primary model are provided in Supplemental Digital Content 1, Appendix S1.4, https://links.lww.com/AA/D549.²³

Primary and Secondary Analyses

In the primary analysis, we tested the hypothesis that a MAP of <55 mm Hg was associated with early perioperative ischemic stroke. In secondary analyses, we tested the effect of a MAP decrease by >30% from baseline on the perioperative stroke. We tested the hypothesis that a patient’s a priori risk of perioperative stroke, as defined by the STRAS prediction score, was associated with a higher duration of intraoperative hypotensive minutes. We also investigated whether intraoperative hypotension was associated with a higher risk of ischemic stroke within 1 year after surgery.

Exploratory and Sensitivity Analyses

With an exploratory intent, we examined whether the STRAS prediction score was associated with a higher risk of perioperative stroke (Supplemental Digital Content 1, Appendix S2, https://links.lww.com/AA/D549).

We conducted several sensitivity analyses to confirm the robustness of our primary findings: we evaluated the effect of intraoperative hypotension on perioperative stroke in subgroups of patients with a high a priori risk of stroke, non-emergency/non-trauma surgeries, and patients undergoing general, urological, and vascular surgeries as well as neurosurgery. A potential modifying effect on the association between intraoperative hypotension and perioperative stroke by age, obesity, intraoperative vasopressor use, history of arterial hypertension, duration of surgery, workRVUs, and preoperative use of antihypertensive medications was assessed by using interaction term analyses. To account for potential bias arising from the combined nature of the STRAS prediction score, we additionally adjusted the primary model for all predictors of the STRAS prediction score individually. Further sensitivity analyses included alternate variable categorizations, additional confounder adjustments, accounting for provider variability in blood pressure management, and multiple imputation of missing covariate data as well as analyses related to loss of follow-up. Details on all sensitivity analyses are described in Supplemental Digital Content 1, Appendix S3, https://links.lww.com/AA/D549.

Power and Sample Size

We conducted a power calculation before the start of data analysis. In our cohort, 160,109 patients experienced intraoperative hypotension and 198,282 did not. Based on a previous publication that investigated the effects of beta blockers on perioperative stroke incidence in patients undergoing noncardiac surgery and observed a 1.0% incidence of stroke in patients who received preoperative metoprolol and a 0.5% incidence in patients who received a placebo, a 0.5% difference in stroke risk between groups was expected (effect size, 0.167).²² Assuming a 2-tailed α of 0.01 and using a 2-sample χ² test, our cohort provided a power of >0.90 to detect a significant association between intraoperative hypotension and perioperative ischemic stroke.

Statistical Analyses

Covariates demonstrating a linear relationship with the primary outcome were included as continuous variables, and variables with a nonlinear relationship were categorized into quintiles or clinically relevant categories (Supplemental Digital Content 1, Appendix S1.4, https://links.lww.com/AA/D549). The complete-case method was used to address missing data in the primary analysis. Missing data were present in 0.7% of overall cases included in this study (0.2% missing STRAS prediction score and 0.5% missing workRVUs). To address potential bias caused by excluding cases with missing data, we conducted multiple imputations by chained equations (5 imputations) in a sensitivity analysis. We used a multivariable logistic regression adjusted for the aforementioned covariates in the primary and secondary analyses to assess the effect of intraoperative hypotension on stroke. A negative binomial regression model was used to evaluate the association between the STRAS prediction score and intraoperative hypotensive minutes as a count variable. For effect modification analyses, interaction terms were included separately in the primary regression model. Results are reported as unadjusted or adjusted odds ratios (OR) for logistic regression models and unadjusted or adjusted incidence rate ratios (IRR) for negative binomial regression models. A P value of <0.05 was considered statistically significant. Statistical analyses were performed using Stata (version 15, StataCorp LLC, College Station, TX).

RESULTS

Study Cohort and Characteristics

Table 1. - Characteristics of the Study Cohort by Occurrence of a Mean Arterial Blood Pressure of <55 mm Hg

Patient outcomes and baseline characteristics	No hypotension (n = 198,282)	Short duration of hypotension (n = 150,429)	Prolonged duration of hypotension (n = 9680)
7-d ischemic stroke, n (%)	742 (0.4)	731 (0.5)	80 (0.8)
1-y ischemic stroke, n (%)	2810 (1.4)	2524 (1.7)	205 (2.1)
STRAS prediction score, median (IQR)	5 (3–10)	6 (4–11)	6 (2–11)
Age, mean ± SD	53.0 ± 16.1	55.5 ± 17.1	51.9 ± 19.6
Female, n (%)	101,098 (51)	91,937 (61.1)	5880 (60.7)
Body mass index (kg/m2), mean ± SD	28.5 ± 6.6	28.2 ± 7.1	27.3 ± 6.7
ASA ≥3, n (%)	62,460 (31.5)	58,841 (39.1)	4216 (43.6)
Charlson comorbidity index, median (IQR)	1 (0–3)	1 (0–3)	2 (0–4)
Underlying disease, n (%)
Ischemic stroke	5461 (2.8)	4822 (3.2)	341 (3.5)
Carotid stenosis	3694 (3.9)	3396 (5.7)	279 (10.2)
Patent foramen ovale	1846 (0.9)	1481 (1.0)	105 (1.1)
Valvular heart disease	12,736 (13.9)	9875 (16.6)	546 (19.9)
Coronary artery disease	25,896 (13.1)	22,661 (15.1)	1566 (16.2)
Atrial fibrillation	13,385 (6.8)	12,399 (8.2)	872 (9.0)
Hypertension	77,697 (39.2)	63,539 (42.2)	3644 (37.6)
Diabetes	13,622 (14.8)	9966 (16.7)	568 (20.7)
Dyslipidemia	32,487 (35.3)	23,350 (39.2)	1107 (40.3)
Smoking	20,841 (22.7)	15,336 (25.7)	833 (30.3)
Emergency surgery, n (%)	10,100 (5.1)	9722 (6.5)	941 (9.7)
Duration of surgery (min), mean ± SD	138 ± 90	173 ± 114	272 ± 168
Work relative value units, median (IQR)	11.4 (6.6–17.7)	14.7 (8.0–21.6)	17.6 (10.5–26.4)

Abbreviations: ASA, American Society of Anesthesiologists; IQR, interquartile range; SD, standard deviation; STRAS, STRoke After Surgery.

Figure 1.: Study flow. *Multiple exclusions may apply. ASA indicates American Society of Anesthesiologists; STRAS, STRoke After Surgery; workRVU, work relative value units.

A total of 156,301 and 204,857 patients underwent noncardiac surgery under general anesthesia during the study period at MGH and BIDMC, respectively. After application of exclusion criteria and exclusion of cases with missing data, the primary cohort consisted of 358,391 patients (Figure 1). A MAP of <55 mm Hg occurred in 44.7% (n = 160,109) of patients, of whom 42% (n = 150,429) experienced a short duration of intraoperative hypotension and 2.7% (n = 9680) a prolonged duration of intraoperative hypotension. In a subgroup of patients with available data (n = 204,067), a decrease in MAP by >30% from baseline occurred in 78.3% (n = 159,732) of patients, of whom 49.4% (n = 100,782) experienced a short duration of intraoperative hypotension and 28.9% (n = 58,950) a prolonged duration of intraoperative hypotension. Table 1 shows characteristics of the primary cohort by the occurrence of a MAP of <55 mm Hg.

Primary and Secondary Analyses

In our cohort, the incidence of early perioperative stroke varied from 0.1% (n = 67/60,316) in general surgery to 2.3% (n = 328/14,369) in vascular surgery. A total of 1553 (0.4%) patients presented with perioperative ischemic stroke within 7 days after surgery. Using a MAP <55 mm Hg as the primary exposure, early perioperative stroke occurred in 742 (0.4%) patients without intraoperative hypotension, 731 (0.5%) patients with a short duration of intraoperative hypotension, and 80 (0.8%) patients who experienced a prolonged duration of intraoperative hypotension. In a crude analysis, the odds of perioperative ischemic stroke within 7 days after surgery were significantly increased for patients with both a short and a prolonged duration of intraoperative hypotension compared to no intraoperative hypotension (Table 2). In an adjusted analysis, no association between a short (OR_adj, 0.95; 95% confidence interval [CI], 0.85–1.07; P = .417) or prolonged (OR_adj, 1.18; 95% CI, 0.91–1.55; P = .220) duration of intraoperative hypotension and perioperative ischemic stroke within 7 days after surgery was found.

Table 2. - Summary of Primary and Secondary Analyses

	Incidence n (%)	Unadjusted odds ratio (95% CI)	Adjusted odds ratio (95% CI)
Primary analysis
Association between mean arterial blood pressure <55 mm Hg and early perioperative stroke (n = 358,391)
No hypotension (n = 198,282)	742 (0.4)	1	1
Short (<15 min) duration of hypotension (n = 150,429)	731 (0.5)	1.30 (1.17–1.44)	0.95 (0.85–1.07)
		P <.001	P = .417
Prolonged (≥15 min) duration of hypotension (n = 9680)	80 (0.8)	2.22 (1.76–2.80)	1.18 (0.91–1.55)
		P <.001	P = .220
Secondary analysis
Association between mean arterial blood pressure decrease > 30% from baseline and early perioperative stroke (n = 204,067)
No hypotension (n = 44,335)	45 (0.1)	1	1
Short (<15 min) duration of hypotension (n = 100,782)	99 (0.1)	0.97 (0.68–1.38)	0.97 (0.67–1.42)
		P = 0.855	P = 0.883
Prolonged (≥15 min) duration of hypotension (n = 58,950)	119 (0.2)	1.99 (1.41–2.81)	1.30 (0.89–1.90)
		P <0.001	P = 0.176

Abbreviation: CI, confidence interval.

Figure 2.: Association among intraoperative hypotension, perioperative stroke, and preoperative STRAS. A, Odds of perioperative stroke as a function of the duration of intraoperative hypotension defined as an absolute MAP <55 mm Hg (left) and a relative MAP decrease >30% from baseline (right); dots/ squares: OR_adj, error bars: 95% confidence intervals. B, Predicted duration of intraoperative hypotension (MAP <55 mm Hg) as a function of the preoperatively calculated stroke risk. A higher preoperative stroke prediction score (STRAS) value was associated with higher predicted intraoperative duration of hypotension (*P < .001). Dots: mean predicted duration of intraoperative hypotension; gray area: 95% confidence interval. C, Probability of perioperative stroke as a function of the preoperatively calculated stroke risk in percent. A higher preoperative STRAS prediction score was associated with higher probability of stroke (**P < .001). Dots: mean predicted probability of stroke; gray area: 95% confidence intervals. MAP indicates mean arterial blood pressure; OR_adj, adjusted odds ratio; STRAS, STRoke After Surgery.

In secondary analyses, we found no association between a decrease in MAP by >30% from baseline and perioperative ischemic stroke within 7 days after surgery (Figure 2A; Table 2). A higher a priori stroke risk quantified based on preoperatively available risk factors (STRAS prediction score) was associated with a longer duration of a MAP <55 mm Hg (IRR_adj, 1.04; 95% CI, 1.04–1.05; P < .001 per 5 points of the STRAS prediction score; Figure 2B). There was no association between arterial hypotension and ischemic stroke within 1 year after surgery (MAP <55 mm Hg of a short (OR_adj, 0.97; 95% CI, 0.91–1.03; P = .255) and prolonged (OR_adj, 1.04; 95% CI, 0.88–1.22; P = .671) duration.

Exploratory and Sensitivity Analyses

In our exploratory analysis, a higher a priori stroke risk quantified based on preoperatively available risk factors (STRAS prediction score) was associated with a higher odds of perioperative stroke (OR_adj, 1.27; 95% CI, 1.26–1.28; P < .001; Figure 2C). Details are provided in Supplemental Digital Content 1, Appendix S2, https://links.lww.com/AA/D549.

Our primary findings remained robust across several sensitivity and subgroup analyses. In a subgroup of patients with a high a priori risk of stroke; non-emergency/non-trauma surgeries; and patients undergoing general, vascular, and urological surgeries as well neurosurgery, intraoperative hypotension was not an independent predictor of perioperative stroke. No significant effect modifiers were observed. When using the individual predictors of the STRAS prediction score as covariates in an adjusted logistic regression model, intraoperative hypotension was not found to increase the risk of perioperative stroke. The primary findings were confirmed after imputation of missing data and all other sensitivity analyses, as detailed in Supplemental Digital Content 1, Appendix S3, https://links.lww.com/AA/D549.

DISCUSSION

In this large multicenter retrospective cohort study, we found that intraoperative hypotension was not an independent predictor of perioperative stroke. However, a high preoperative stroke risk (based on the STRAS prediction score) was associated with increased odds of intraoperative hypotension and increased odds of early perioperative stroke.

Approximately half of the patients had intraoperative hypotension, which continued for a prolonged period in 2.7% of patients. Our data show that a high a priori risk of stroke (STRAS) was a predictor of intraoperative hypotension, highlighting the fact that preoperative risk factors of intraoperative hypotension and adverse outcomes often coexist in the same patient.^24,25 For example, the presence of cardiovascular or peripheral vascular disease puts a patient at a higher risk of both intraoperative hypotension and perioperative stroke.^3,14,26 There is equivocal evidence on the effect of intraoperative hypotension on perioperative stroke. While some studies have suggested a directly harmful effect of intraoperative hypotension on cerebral blood flow (CBF),^12,13 others have proposed rather indirect effects coming into play if cerebral autoregulation is compromised due to underlying conditions.^14–16,27

Our data show that adjusted for a patients' risk of perioperative stroke according to the stroke prediction score we have used, intraoperative hypotension in the range studied does not increase perioperative stroke risk, emphasizing that CBF and, hence, perfusion of the brain are subject to effective autoregulation across a wide range of arterial blood pressures.^28–30 The current mainstay treatment in stroke prevention in the general population is decreasing a patient’s elevated systolic blood pressure.³¹ Post hoc analyses from 2 randomized controlled trials have suggested that aggressively lowering blood pressure even to achieve MAPs below 60 mm Hg do not increase stroke risk.³² Our findings are in agreement with these data to the extent that intraoperative hypotension was not associated with increased odds of early perioperative ischemic stroke. These findings indicate that the long-held belief that lowering mean arterial blood pressure below 60 mm Hg increases stroke risk may be incorrect, and that MAP values below this level may be safe in terms of perioperative stroke risk. However, clinicians should continue to avoid significant hypotensive episodes during noncardiac surgery if not to decrease the risk of stroke in patients, then to minimize the risk of other neurological complications.

There is currently no agreement on an optimal MAP range that is generalizable to patients undergoing surgery. A variety of different thresholds for defining a low intraoperative MAP or significant decreases in MAP from baseline have been examined.^13,27,33 Studies using CBF monitoring (eg, near-infrared spectroscopy or transcranial Doppler ultrasound) showed that an interindividual variability exists (range of MAP, 40–90 mm Hg) in the MAP threshold required to maintain constant CBF.³⁴ Thus, MAP thresholds of 55 mm Hg as used in our study were below the lower limit of cerebral autoregulation and may result in hypoperfusion in patients with an elevated lower limit of CBF autoregulation. An individualized MAP target may be selected based on real-time CBF measurements to reduce neurological complications and the risk of postoperative organ dysfunction.^35–37 However, this real-time monitoring of cerebral autoregulation is technically challenging and cannot be used during routine anesthesia care, since it requires the availability of experts in CBF measurement in the operating room.

Limitations arise from the retrospective design of the presented study. Although the large cohort size allowed for a wide range of adjustment for potential confounding variables and the robustness of our results was confirmed in several sensitivity analyses, residual unidentified confounding cannot be entirely excluded. Second, the outcome “diagnosed stroke” in this study does not capture “silent strokes", such as cases in which no clinical symptoms of stroke exist (eg, aphasia or deficits in motor or sensory function), although there would have been evidence of an ischemic brain injury if magnetic resonance imaging had been performed.^38,39 Third, the real incidence of diagnosed stroke within 1 year after surgery was likely higher than the incidence in our report due to a potential loss of follow-up to some patients. However, sensitivity analysis addressing a potential loss of follow-up confirmed the robustness of our findings. Finally, our results may lack generalizability, since we analyzed data from a limited geographical region. However, we present a multicenter study with hospital sites performing a large variety of procedures. We, therefore, believe that our findings provide clinically relevant information and could guide clinicians on the management of intraoperative hypotension.

In conclusion, a decrease in mean arterial blood pressure to intraoperative values of an MAP of <55 mm Hg during noncardiac surgery is not an independent risk factor of perioperative ischemic stroke. These findings emphasize the importance of CBF autoregulation in the perioperative setting.

DISCLOSURES

Name: Karuna Wongtangman, MD.

Contribution: This author helped conceptualize the article, perform literature review, analyze data, write the initial draft of the manuscript, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Luca J. Wachtendorf, cand.med.

Contribution: This author helped conceptualize the article, analyze the data, write the initial draft of the manuscript, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Michael Blank, cand.med.

Contribution: This author helped analyze the data, review the manuscript, edit and modify the drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Stephanie D. Grabitz, MD.

Contribution: This author helped give suggestions on data analysis, review the manuscript, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Felix C. Linhardt, cand.med.

Contribution: This author helped analyze data, review the manuscript, edit and modify drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Omid Azimaraghi, MD.

Contribution: This author helped give suggestions on data analysis, perform literature review, review the manuscript, edit and modify drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Dana Raub, cand.med.

Contribution: This author helped give suggestions on data analysis, perform literature review, review the manuscript, edit and modify drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Stephanie Pham, BS.

Contribution: This author helped give suggestions on data analysis, review the manuscript, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Samir M. Kendale, MD.

Contribution: This author helped give suggestions on data analysis, review the manuscript, edit and modify drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Ying H. Low, MD.

Contribution: This author helped give suggestions on data analysis, perform literature review, review the manuscript, edit and modify drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

Name: Timothy T. Houle, PhD.

Contribution: This author helped give suggestions on data analysis, supervise the study, review the manuscript, and read and approve the final manuscript.

Conflicts of Interest: T. T. Houle reports grants from National Institute of Neurological Disorders and Stroke (PI), grants from National Institute of General Medical Sciences, personal fees from Headache, personal fees from Anesthesiology, and personal fees from Cephalalgia outside the submitted work.

Name: Matthias Eikermann, MD, PhD.

Contribution: This author helped conceptualize the article, give suggestions on data analysis, supervise the study, review the manuscript, edit and modify drafts, and read and approve the final manuscript. He is the guarantor of the study who takes responsibility for all parts from conceptualization to publication.

Conflicts of Interest: M. Eikermann received an unrestricted grant from Jeffrey and Judith Buzen (222302) and a grant from Merck & Co for work not related to this study. He holds equity of Calabash Bioscience Inc. He is an Associate Editor of the British Journal of Anaesthesia; no other relationships or activities could have influenced the submitted work.

Name: Richard J. Pollard, MD.

Contribution: This author helped give suggestions on data analysis, perform literature review, supervise the study, review the manuscript, edit and modify the drafts, and read and approve the final manuscript.

Conflicts of Interest: None.

This manuscript was handled by: Oluwaseun Johnson-Akeju, MD, MMSc.